Why when watching cars travelling at high speeds do the wheels appear to be going backwards?

I’m guessing you mainly see it in movies or, less commonly, videos. It’s a stroboscopic effect related to the fact that moving pictures are actually a rapid sequence of still frames. If the frame rate is synched to the rotation of a wheel (or to some repeating feature like the spokes of a wagon wheel) then the wheel will appear to not move, or to move slowly forwards or backwards.

Using normal vision directly, motion appears as a blur. You need a discrete frame rate to produce this effect.

as @gasman has pointed out its the stroboscopic effect. but if you want to experiment with this in real life with your own two eyes there is a way, all you need is a strobe light and a fan or other spinning object. if you set the strobe to match the RPM of the object you can actually make it look like it is stopped, and if you then speed the strobe up or slow it down you can make it look like it is spinning backwards or forwards.

Are you saying the turn of the wheel is turning faster than…what is it turning faster than? I get the strobe light example.

The stroboscopic effect, as mentioned above, also works for normal vision. The part of your visual field capable of showing fine detail is only about the size of your thumb at arms length. Your eyes go through saccades to compensate and form a larger mental image. So as your eyes dart around gathering information, the wheel has moved. An experiment to confirm this is to stare intently at the centre of the wheel, making sure you focus enough to suppress your saccades. After a few seconds the effect should disappear, as all but the centre of the wheel will be blurred since you have no detailed information about it.

Just a quick note; many lights flicker anyways. Lights powered by AC power rather than batteries flicker quite a bit, as do the headlights on many motorcycles. Flashlights generally do not flicker since they are powered directly from a DC source at a constant voltage that doesn’t change polarity many times a second like AC does.

This can be dangerous in that you can have an electric motor lit by AC lighting look like it’s standing still when it’s actually spinning at an RPM that coincides with the flicker of the lights. If you work the type of places I have/do, that could cost you at least a few stitches…. if it doesn’t kill you.

@jerv: Good point about artificial light, especially fluorescent tubes. I don’t think there’s any discernible flicker from AC-powered incandescent filaments (not without sensitive instruments, at any rate). TVs and various CRT displays definitely have flicker that can induce strobe effects. LEDs driven by variable duty-cycle square waves (for dimming) essentially are strobe lights, though usually at much higher flicker rates than the frame rate of movies, which varies from 16 to 60 Hz. (24 at the movie theater).

Motorcycle lights flicker at relatively low rates to increase their visibility—but I doubt fast enough to produce stroboscopic effects. Then again, I don’t ride them & never looked at their reflections. You’d have to have a fairly high-grade amount of flicker before an electric motor would actually appear to stand still. @jerv, are you saying this poses an actual danger to people?

I’ve been playing with strobe lights since I was a kid—even did a high school report on them. I had access to (& later purchased) what’s now a long-obsolete piece of laboratory equipment called a Strobotac—a neon-light capable of flickering at a precise user-set rate up to around 30 or 40 kHz! (The typical strobe that you buy today at Radio Shack or Spencer’s for dance set-ups barely flashes above the threshold of flicker—around 16 / sec, based on human retinal persistence of vision.)

You can make a simple stroboscope yourself, btw, from a spinning disk with slots cut into it at regular intervals (I used to use unwanted phonograph records, but cardboard will suffice). Spin by hand. View through a fixed slit that lines up with the spinning ones. It’s a slow-mo time machine for any rotational or repetitive motion, no matter how fast.

@FireMadeFlesh: I understand saccades, but I’m not sure what you mean. When I stare at the center of a rapidly rotating object, I already see a blur everywhere else. I find that the rapid, darting saccade motion can temporarily give me a sharp view of detail otherwsie blurry from motion. Thus if I saccade to the left while looking at the bottom of a spoked wheel rotating clockwise, I can see individual spokes. Just a brief glimpse, however—no way to sustain the view as with a strobe. Photographers employ this effect (rapid panning) to catch a sharp image of a fast-moving object such as a passing race car.

@gasman Individual tolerances vary, and I am glad to be of an age where I am starting to lose my high-range hearing since it means I no longer wince at sounds that others cannot hear. However, I cannot use a CRT monitor with a refresh rate below 70Hz, and if I’ve had too much sugar and/or caffeine, even household incandescent lights become a little bit of a nuisance. (I can’t actually see it flicker, but I notice it “wavering” enough to be distracting.)

I have used many a Stobotac in my day, in part since the Navy will use “obsolete” equipment if it is reliable. Overhead lighting can sometimes become, for lack of a better term, an uncalibrated, fixed-rate Strobotac. It’s uncommon since most AC motors spin slightly slower than the 1800 RPM than you would expect from a 60Hz input (usually closer to 1700–1750), but most =/= all. I have seen the variable speed chucks/spindles on CNC equipment appear to stand still; fortunately they are loud enough than you can hear the motors even with earplugs in.

Motorcycle lights flicker at a much lower rate and really can’t do a stroboscopic effect, especially since they are more of a “dim-bright-dim-bright” flicker than “on-of-on-off”.

@gasman The stroboscopic effect works when there is a flashing light, but the effect of seeing a car’s wheel apparently turning backwards can also be seen in broad daylight. This is because the speed of rotation of the wheel is such that it almost matches the rhythm of your saccades. Your eyes will dart around within your focus on (for instance) the top of the wheel, trying to focus on the spokes. It is much easier at some speeds to go against the direction of motion, and so it appears to be moving backwards because your eyes are moving backwards to keep up. You can show that the wheel is rotating forwards if you force particularly rapid saccades, because then you can focus on a spoke going forwards for that brief moment. My point about staring at the centre of the wheel is simply that if you suppress your saccades the effect disappears.